Stillbirths account for half of all perinatal mortality, but the underlying cause of a significant portion of the cases remains unknown. We set out to test the potential role and extent of microbial infection in stillbirth through a case-control analysis of fetal cord blood collected from the multisite Stillbirth Collaborative Research Network. Cases (n = 60) were defined as stillbirths at .20 weeks of gestation, and controls (n = 176) were live births. The bacterial presence, abundance, and composition were analyzed by endpoint PCR of full-length 16S rRNA and the V4 amplicon sequence variants (ASVs). The results demonstrate that bacterial prevalence and abundance were both significantly increased in stillbirth, even after adjusting for maternal age, race, body mass index, number of pregnancies, gestational age, and multiple gestations. Composition of bacterial communities in the cord blood also differed significantly. Using a group of 25 ASVs differentially abundant between the two groups, a Random Forest classification model achieved an accuracy score of 0.76 differentiating stillbirth and live birth, with Group B Streptococcus as the most enriched species in stillbirth. Positive PCR was also significantly associated with early preterm birth. A group of oral anaerobes, including Actinomyces, Campylobacter, Fusobacterium, Peptostreptococcus, Porphyromonas, and Prevotella, were enriched in live early preterm birth, suggesting possible oral origin of infection. Our ASV-based microbiome analysis revealed specific candidate pathogens associated with infections in stillbirth and early preterm birth. The cord blood microbial signatures may be markers of adverse pregnancy outcomes. Our study will help identify possible mechanism of infection and improve our ability to prevent stillbirth and early preterm birth.